Two Methods Of Robust Design Optimization And Research On Integrated Design Of Performance Robustness And Manufacturing Accuracy

Uncertainties are inevitable in the design process of underwater structure due to the lack of knowledge and the changfully operation environment.Continuously transmitted and accumulated,the uncertainties may cause some performance indicators to change and shift,even lead failure.Therefore,it is necessary to consider the influence of uncertainties in the robust design optimization of underwater structures.However,conventional nested optimization methods are often time-consuming.In recent years,some methods based on the surrogate model have greatly improved the efficiency in the design of underwater structures,but there is always an error between the response estimated by the surrogate model and practical response,namely,the error of surrogate model.In addition,the manufacturing accuracy in the manufacturing process has a great influence on the performance robustness of underwater structures.In addition,manufacture error also has great influence on the performance and robustness of underwater structures.Therefore,it has become urgent problems that how to solve the robust design optimization of underwater structures efficiently and make integrated design of performance robustness and manufacturing accuracy,considering both the uncertainty and the error of surrogate model.In order to solve the problems above,the specific research works are as follows:(1)Firstly,a new robust design optimization method based on previous optimization knowledge(PK-RO)is proposed to solve the problem with interval uncertainty.In the proposed method,the individuals’ response values that have been accurately calculated within the critical distance are used to approximately predict the response values of other individuals and to evaluate the robustness indexes accordingly.Besides,the information of the accurate individuals’ response values,which is gradually expanded in the evolutionary procedure,is used to selectively re-evaluate the robustness of some past individuals,and the critical distance is adaptively reduced based on the misjudgment rate of the robustness.Three numerical examples are tested to demonstrate the applicability of the proposed algorithm.The results show that the proposed algorithm saves more than 94% of the computational resources while the estimated error is less than 2.5%.(2)Secondly,a new sequential updating surrogate assisted robust optimization method(SU-RO)is proposed to improve the efficiency of comupation,which both integrates the information of robust optimal and the accuracy of surrogate model.Kriging surrogate model is used to replace the time-consuming finite element calculation,and the Monte-Carlo simulation based on surrogate model is used to evaluate the robustness index.Then,updating strategies according to the robust optimal and the accuracy of objective function and constraint function surrogate model are proposed respectively to ensure the accuracy and robustness of the solution.Results of four numerical examples demonstate that the proposed SU-RO method has superior to other robust optimization methods,because it greatly reduces the computational resources and obtains the better robust solution.(3)Thirdly,a mathematical model for integrated design of performance robustness and manufacturing accuracy is proposed to obtain both robust solution and manufacturing accuracy(design tolerance).The modified SU-RO method is applied to the integrated design of similified inner pressure tank and metamaterial vibration isolator.Under the condition of fixed performance and robustness requirements,the optimal robust solution and its corresponding tolerance radius are obtained respectively.With the relaxation of performance or robustness constaints,the allowable maximum manufacture error gradually increases,which means the required manufacturing accuracy becomes lower.This thesis provides new methods and guidance for the robust design optimization and manufacturing accuracy control of typical underwater structures.